Enhancing patient safety through the quality assured use of a low-tech video interpreting system to overcome language barriers in healthcare settings

The negative consequences of language barriers in the public healthcare system are apparent. A patient’s command of language determines help seeking [5], ability to orientate within the healthcare system, and thus to gain access to medical resources [6‐8]. This results in unequal access to healthcare, low patient satisfaction [9‐11] and makes the quality of medical treatment highly dependent upon linguistic capabilities [12‐16]. The situation further represents a dangerous risk and challenge for quality assurance and patient safety [17]. The conversation between patient and healthcare professional suffers most from language barriers.

Betancourt et al. [18] cited the three most common sources of adverse events for foreign language patients: First, the use of unqualified language mediators (relatives and staff, etc.) [19‐21], second, the influence of cultural beliefs and traditions on care [22] and third, the misconception of employees to master the foreign language well enough (“get by”) [20, 22].

The negative impact of communication barriers on the importance of shared decision-making in patient safety has been recently demonstrated by Barton in a study with a large sample of rheumatoid arthritis patients [23]. Considering the importance of the doctor-patient conversation as a diagnostic and therapy tool [24‐26], the ramifications can be serious, especially in highly dangerous outcomes [27]. It is harder for patients who do not speak the local language to provide adequate informed consent, make treatment decisions, follow further instructions, and manage follow-up dates [28]. Especially in emergency situations, lay interpreters such as hospital employees and family members are employed. This carries risks, including confidentiality, misunderstandings with possibly severe medical consequences and an inadequate emotional load on children and other ad hoc interpreters [1, 4, 29, 30].

The medical consultation is the basis for the formation of informed consent. A language barrier can therefore hinder gaining informed consent which is not only a prerequisite for a medical indication, but also for lawful treatment [31, 32]. The right to informed consent as a foundation for proper treatment is anchored in Article 3 of the World Health Organization (WHO) Declaration on the Promotion of Patients’ Rights in Europe [33], Article 6 of the Universal Declaration, the Biomedicine Convention of the Council of Europe on Bioethics and Human Rights [34], as well as in Article 8 of the European Convention of Human Rights (ECHR) [35] and Article 3 of the Charter of Fundamental Rights of the European Union (CFR) [36]. There is, compared to other countries such as the USA, however, no explicit right to medical consultation in one’s native language, although the consultation must allow the patient to make an informed, self-determined decision. The burden of proof that the patient has understood the information communicated lies with the treating healthcare professional and/or the hospital [16]. Where language barriers prevent informed consent, the healthcare professional is not obliged to conclude a treatment contract and risks legal consequences. The only exceptions to this are medical emergencies [37].

Diagnosing and treating patients with language barriers involves higher costs. This should not guide safety considerations, but costs can be significantly reduced by professional interpreting services. After evaluating 4146 hospital stays in an American emergency room Hampers and McNulty concluded that drawing on the services of a professional interpreter represented the lowest cost burden and resulted in shorter hospital stays, reduced administration of unnecessary medication, and appropriate treatment [38]. Remote interpretation methods can reduce other inefficiencies, such as traveling and scheduling [39]. Furthermore, interpreting visual clues such as gestures, pointing, or facial expressions can help overcome cultural barriers, which makes it the primary reason why patients prefer video interpreting to telephone interpreting in the few studies published [40, 41].

A two-step mixed methods design was used to collect and examine the relevant research data focusing on prior experience, satisfaction, and experiences during the pilot test. Quantitative questionnaires were completed by the healthcare professionals and the VI software recorded relevant data, such as the time stamp, call duration and language used. This was followed by qualitative semi-structured interviews conducted with the same healthcare professionals. Over the course of the project questionnaires were also sent to the patients, and semi-structured interviews were conducted with the interpreters: the results of these data will not be included in this article, which limits its focus to the healthcare professionals.

The semi-structured interviews were processed using common theoretical coding [44] and qualitative content analysis [45]. This ensured that the transcribed material was methodically controlled, analyzed, and coded. Codification was first open, then selective, with the aim of defining core categories from the previously processed code [45].

Codes were used as a means of breaking down and then reassessing the text. By developing and modifying the codes, a new order emerged in the interviews and this in turn facilitated the processing. It became easier to compare the empiric material with the activity patterns and conditions [44]. Qualitative content analysis method allowed activity patterns and social realities to be visualized.

The descriptive statistics analysis tool was applied to the quantitatively collected data. It served to calculate different parameters predetermined by the research questions, displaying them as graphics and tables. Descriptive statistical parameters within a questionnaire describe only what applies to the respondents of the questionnaires, which is why they cannot provide statements regarding the whole population. Thus, this analysis tool shows the structure of observations and the representation of this structure in form of tables and graphics [46].

Finally, once the qualitative and quantitative materials had been analysed separately, a mixed analysis was undertaken. The meta-inference, i.e. the comparative viewing of the inferences of both strands, composes the “crucial gain of a mixed-methods study, which does not only contain separate method strands, but also method strands that are analysed specifically in relation to each other” [47].

Of the respondents 78% indicated that conversation in the native language was only possible to a limited extent, while 7% noted having found themselves in situations in which conversation with patients were so problematic that the consultation had to be aborted. In such cases the only possible recourse was to use an acute solution strategy, such as locating a staff member who speaks the relevant language or inviting the patient to return for a new appointment with the request that they bring someone as a lay interpreter. Both solutions can lead to unsatisfactory results, set patient safety at risk and conflict with medical ethics as in the case of children used as interpreters.

The majority of used lay interpreters were not employed by the medical establishment for that purpose [25]. Most are family members, friends of the patient, or employees of the medical establishment. The involvement of a child for interpretation can lead to emotional burden and emotional problems for the child [48] and threatens the communication process [4, 19]. As we demonstrated in the pilot project, healthcare professionals often face conflicting interests. They need to maintain the workflow of the medical establishment with no interruptions and delays, whilst sharing the broad consensus among healthcare professionals that this solution is one of the least appropriate and least ethical.

Another solution strategy which creates a clear ethical conflict is the use of other patients as lay interpreters. Of the healthcare professionals 42% indicated that they have used this solution strategy despite wanting to avoid associated problems.

Other common practise strategies are gestures and postures (60%), use of translated information materials (34%), drawings, and writing down information (Fig. 3).

Considering the problematic aspects of using non-professional language mediators, it was unsurprising that healthcare professionals tended to answer the question on satisfaction with previous strategies in the negative (Fig. 4).

As the satisfaction level is skewed towards dissatisfaction, there is potential for improvement and room for a different solution strategy. To what extent VI as a solution strategy can fill this gap is analyzed in the second part of this study. The following section addresses the use of this solution strategy and its satisfaction level.

The healthcare professionals’ personal impressions of key aspects of the VI solution were collated. They identified the following as crucial: rapid availability, data protection compliance, the ability to see the interpreter, the absence of potential external influence resulting from personal relationships, the user-friendly technique, the legal certainty, the absence of the requirement for personal presence, and cost savings. This is congruent with the problems experienced when treating nonnative speakers noted above (Fig. 5).

Few users rated it as mediocre (5%), unsatisfactory (3%) or bad (2%). One respondent (2%) did not answer.

First, we documented the low satisfaction with previous methods that are based on inadequate strategies.

Second, we focused on VI with professional interpreters. This part of the study demonstrated that at least on the level of professional interpretation this strategy had a major impact. It became clear that the intervention, quickly and easily accessing professional interpreters through video calls, supported the work of healthcare professionals and relieved staff of additional workload.

Furthermore, the lack of physical presence did not prove a hindrance to language mediation, underscoring the logistical and cost advantages of VI. The study also suggested that, while the duration of the consultations decreased, practitioners perceived a subjective increase in quality.

Introducing this tool involved changes to existing routines and structures. As a result, active involvement of the management, and particularly quality management, are prerequisites for successful long-term implementation. Some pre-existing management issues could hinder the successful implementation of VI. A difference in satisfaction levels could also be observed correlating with the logistical and technical competence of the endpoint.

Strengths of the study lie in the mixed methods approach, and the inclusion of different medical settings within the sample, to accurately interpret and attribute potentially different outcomes. Only around 30 healthcare professionals attended the original preparation workshops. This was remedied by intense and personal introductions given by the interpreters during first use.

Resource limitations meant that even during the study design the intended study duration of 1 year would not be feasible. As end-users needed to first familiarize themselves with the method, its overall use was lower than expected. Another obstacle was the limited selection of languages on offer, as many more were needed.

It was not possible to monitor technical implementation at endpoints. As a result, successful implementation was dependent on the local technical services, and on support from the commercial company supplying the solution. Therefore, potential influence of researchers on the results was reduced to a minimum.

Limited availability of certain language interpreters, such as Turkish, who met the required professional and academic criteria was problematic. This could be an issue for large-scale application and academic training for languages such as Urdu may be non-existent.

This study focused on the healthcare professionals’ view, and has allowed a diversified and specific analysis of data and influencing factors.

Although healthcare professionals in this study noted that language barriers often occur and have stated their willingness to use VI, the long-term impact on healthcare efficiency, satisfaction and economy needs to be examined.

According to the WHO, the right to language mediation is embedded within the right to informed consent [49], and to equal access to treatment.

As the pilot project in this study shows, once interpretation units are set up and healthcare professionals familiarized themselves with this solution, they are willing to use the technology. User friendliness and easy access need to be considered. This can be expected to benefit all aspects of patient safety, quality of medical care, and equal access to healthcare for patients with different language backgrounds.